Access to Research at National University of Ireland, Galway
Research at National University of Ireland, GalwayNot a member yet
15712 research outputs found
Sort by
Is youth mentoring beneficial for child and adolescent mental health service users? A multi-stakeholder perspective
No full textBackground
Internationally, the number of young people who experience significant mental health issues is increasing. It is argued that flexible, community-based initiatives can help support mental health services to address the mental health needs of young people. Youth mentoring is a community-based model, which has been found to act as a supportive resource for vulnerable and at-risk youth.
Objective
This research examines the benefits and challenges associated with the provision of mentoring within a youth mental health context. The study explores the rationale guiding the partnership between a community-based mentoring organisation and child and adolescent mental health services (CAMHS) in Ireland, and identifies key practice considerations.
Method
40 participants involved with the partnership took part in semi-structured qualitative interviews, which were analysed using a thematic analytic approach.
Results
Mentoring was viewed as a means of providing relaxed, informal, friendships that could help the young person to socialise more, strengthen their social skills, and become more integrated into their communities. Positive changes in social and emotional well-being were observed among participating youth. Practical challenges included delays in matching young people with mentors and communication issues.
Conclusions
Findings have relevance for researchers and practitioners interested in the integration of community-based and statutory mental health services, as they indicate that youth mentoring can act as a supportive resource for youth who experience mental health difficulties. Findings also provide insights into the processes that may support/hinder the effectiveness and acceptability of mentoring within a context mental health, which can help inform best practice guidelines.
Similar content being viewed by othersOpen Access funding provided by the IReL Consortium. This research was funded by the Health Service Executive.peer-reviewe
KnowZRel: Common sense knowledge-based zero-shot relationship retrieval for generalised scene graph generation
No full textA scene graph is a key image representation in visual reasoning. The generalisability of Scene Graph Generation (SGG) methods is crucial for reliable reasoning and real-world applicability. However, imbalanced training datasets limit this, underrepresenting meaningful visual relationships. Current SGG methods using external knowledge sources face limitations due to these imbalances or restricted relationship coverage, impacting their reasoning and generalisation capabilities. We propose a novel neurosymbolic approach that integrates data-driven object detection with heterogeneous knowledge graph-based object refinement and zero-shot relationship retrieval, highlighting the loosely coupled synergy between neural and symbolic components. This combination addresses the limitations of imbalanced training datasets in scene graph generation and enables effective prediction of unseen visual relationships. Objects are detected using a region-based deep neural network and refined based on their positional and structural similarity, followed by retrieval of pairwise visual relationships using a heterogeneous knowledge graph. The redundant and irrelevant visual relationships are discarded based on the similarity of relationship labels and node embeddings. Finally, the visual relationships are interlinked to generate the scene graph. The employed heterogeneous knowledge graph combines diverse knowledge sources, offering rich common sense knowledge about objects and their interactions in the world. Our method, evaluated using the benchmark Visual Genome dataset and zero-shot recall (zR@K) metric, shows a 59.96% improvement over existing state-of-the-art methods, highlighting its effectiveness in generalised SGG. The object refinement step effectively improved the object detection performance by 57.1%. Additional evaluation using the GQA dataset confirms the cross-dataset generalisability of our method. We also compared various knowledge sources and embedding models to determine an optimal combination for zero-shot SGG. The source code is available at https://github.com/jaleedkhan/zsrr-sgg.This publication has emanated from research conducted with the financial support of Research Ireland under Grant number 18/CRT/6223 and
12/RC/2289 P2.peer-reviewe
Development of phosphorus recovery technologies for sustainable utilization of livestock manure
No full textPhosphorus (P) recovery from livestock manure is a potential approach to address the depletion of P rock reserves. However, P in manure primarily exists in insoluble solid form, which poses a major challenge to the recovery process since only soluble P can be effectively separated and recovered. Traditional chemical acidification methods, while effective in solubilizing P, are limited by high chemical consumption, elevated costs, and the generation of secondary pollutants, underscoring the need for more sustainable alternatives.
To overcome these limitations, this study investigated biological acidification and electrolytic acidification, focusing on optimizing key reaction conditions and analyzing the underlying mechanisms of P release and acidification. It also investigated the integration of these P release methods with struvite crystallization in terms of recovery performance and product characteristics.
Biological acidification was performed by co-fermenting pig manure (PM) with food waste (FW) under anaerobic conditions. The results showed FW promoted lactic acid production and rapid acidification. As FW increased from 0 to 80%, the concentrations of lactic acid rose from 0.12 ± 0.04 g/L to 11.95 ± 1.37 g/L, with pH decreasing from 7.55 to 4.43. The ratio with FW/PM=1:2 was the optimal condition, which led to the highest dissolved phosphate (PO43--P) concentration of 350.39 ± 8.59 mg/L in 72 h, with a total P release of 74.2 ± 1.8%. Multiple regression analyses established key relationships to predict pH changes in the reactor.
Electrolytic acidification was conducted using an electrochemically mediated phosphorus release (EMR) reactor at different voltages. The results showed over 70% of particulate P was converted to soluble reactive P during the EMR process, involving the dissolution of inorganic P and the conversion of organic P. Increasing the voltage appropriately enhanced ion migration and acidification rates. At 9 V, the release of 92.1 ± 1.3% was obtained in 36 h.
Subsequent P recovery from P-rich supernatants was conducted using struvite crystallization. The fermentation supernatant achieved the highest struvite purity of 91.9 ± 2.2% at pH of 8.0 and an Mg/P ratio of 1.5, with a P recovery of 97.9%. In contrast, the EMR supernatant yielded a struvite purity of 51.7 ± 0.3% under optimal conditions of pH of 9.0 and an Mg/P ratio of 1.75, with a P recovery of 96.3%. Morphological analysis revealed well-defined crystalline struvite from the fermentation supernatant, whereas the product from the EMR supernatant contained irregular aggregates and more impurities, such as brushite and hydroxyapatite.
This research demonstrates the feasibility of sustainable P recovery by integrating acidification and struvite crystallization, offering practical solutions for nutrient recycling. It provides valuable insights for optimizing recovery systems, supporting the circular economy, and mitigating environmental risks associated with manure management.Funded by the China Scholarship Counci
Networks for scaling businesses: Accessing international support webs for critical support resources
No full textEntrepreneurial ventures benefit from breaking out of structural localism to access external supports as they scale their businesses. However, there is limited theory development on how internationally scaling businesses access external supports. By adopting an entrepreneurial network perspective to study the support system for internationally scaling businesses located in Ireland, we identify an international support web accessed through network mechanisms that break out of structural localism. This study contributes to international entrepreneurship by elaborating on network mechanisms for scaling businesses, in contrast with early and rapidly internationalizing ventures and mature multinational enterprises. It also contributes to international connectivity research by detailing micro-processes for cross-border resource flows.peer-reviewe
An grúpa amharclainne na Fánaithe agus a bhfuil le foghlaim óna gcartlann faoin drámaíocht a bhain leo
No full textTá tuairisc anseo ar chartlann a bhaineann leis an ngrúpa amharclainne Na Fánaithe (1987–1993). Cuimsíonn an chartlann foinsí éagsúla: cuimhní pearsanta, gearrthóga ó nuachtáin éagsúla, ábhar clóite eile, clár faisnéise agus píosaí físe. Áitítear go dtugann cartlann na bhFánaithe léargas áirithe ar an amharclannaíocht, ar chúrsaí gníomhaíochais agus ar an luach oidhreachta agus staire a bhaineann le hábhar na cartlainne áirithe seo. Cíortar chomh maith cuid de na deiseanna a sholáthraíonn teicneolaíochtaí digiteacha don chartlannaíocht.peer-reviewe
Game engine based synthetic data generation schemes and convolutional neural networks
No full textThis thesis presents designs and implementations to generate synthetic images using game engines to train CNNs (Convolutional Neural Networks). It also investigates some fundamental properties of CNNs, namely the performance characteristics with different number of target classes; and the training characteristics using our novel learning rate tuning method. To train CNNs for a computer vision problem requires a huge number of annotated images as training data, which is labour intensive and expensive. A part or whole of the training data can be synthesized with a wide variety of methods.
Our first contribution is to synthesize aerial top-down images and thereby, attempt and demonstrate the feasibility of two domain transfers at once, one being synth-to-real (training on synthetic data and predicting on real data), two being front-facing to aerial domain (taking a CNN pretrained on consumer camera images which are primarily front-facing and finetuning/testing that CNN on aerial top-down images). We generated synthetic data for that from a realistic virtual 3D game environment by programmatically flying a (quadrocopter) Robotic Aerial Vehicle (RAV) inside the game and annotating the synthetic images so taken from its camera. We then demonstrated dual domain transfer by detecting aerial-view real-world objects using a CNN trained on our synthetic data.
Our second contribution is the design, development and evaluation of a hybrid synthetic data generation approach that combines the realistic lighting, object placements etc. from the 3D game engine with complex textures and backgrounds sourced from the internet. The network finetuned with synthetic data so collected outperforms the same network finetuned with real data when tested on a challenging dataset called ObjectNet and also sets a state-of-the-art result for any convolutional neural network on ObjectNet.
Our third contribution is an investigative work that delves into performance characteristics of CNNs with increasing number of classes to predict. To that end, we conduct a systematic investigation on three ubiquitous computer vision tasks – image classification, object detection, and semantic segmentation, examining how performance changes with increasing number of class labels, while controlling for variables like CNN architecture and training methodology. We use multiple datasets for each task. We find that in image classification and semantic segmentation, performance decreases with increasing number of classes. Conversely, we discover that performance improves with more classes in object detection. We further explore this observed difference by visualizing and analyzing feature maps in terms of their clustering performance. We conclude that in object detection, the feature map clusters become tighter and better separated as the number of classes increases, leading to an increase in performance.1. European Union’s Horizon 2020 Research and Innovation Programme–Grant Agreement Number 700264 (ROCSAFE);
2. Prof. Michael Madden (PI Research Overheads)
3. . Science Foundation Ireland under Grant Number 12/RC/2289_P2–Insight SFI Centre for Data Analytics (co-funded by the European Regional Development Fund)
Stacking interactions in indomethacin solid-state forms
No full textStacked structures with strong dispersion forces between stack neighbors often lead to anisotropic crystal growth and needlelike morphologies. The crystal structures of a new cocrystal and a molecular salt of indomethacin (IND) are reported: IND·MOA and IND·POBA·0.5H2O (MOA = p-methoxyaniline, POBA = 4-phenoxybenzylamine). In both structures, the IND and coformer molecules/ions are stacked and IND adopts the unusual conformation found in the α-polymorph of pure IND, resulting in a relatively short distance of about 3 Å between the methyl group and the C1′-atom of the chlorophenyl ring. While IND·MOA and IND·POBA·0.5H2O both crystallize as needles like α-IND, the weaker stacking interactions of the coformer in the IND·MOA cocrystal lead to shorter and thicker needles. Amorphous IND prepared by milling recrystallizes to the stable γ-polymorph without the metastable α-form being detected. When IND is milled in the presence of 2.5 wt % MOA, the amorphous phase converts to α-IND. The effect of small amounts of the coformer on the recrystallization route is attributed to a templating effect of the cocrystal formed during milling and/or the facilitation of the conversion to the α-phase conformation.This publication has emanated from research supported in part by a research grant from Science Foundation Ireland (SFI) and is cofunded under the European Regional Development Fund under Grant Number 12/RC/2275-P2. N.F. thanks the Irish Research Council for a Government of Ireland Postgraduate Scholarship (Project ID GOIPG/2023/4906). M.A. acknowledges the King Fahd University of Petroleum and Minerals (KFUPM) for providing facilities for this research. For the purpose of Open Access, the author has applied a CC BY public copyright license to any author accepted manuscript version arising from this submission.peer-reviewe
Beyond the binary: integrating “real-world evidence” with randomized trials in contemporary health care
No full textKey findings
We suggest developing an integrated framework combining randomized controlled trials (RCTs) and real-world evidence (RWE), emphasizing their complementary strengths in health-care research. Currently, this integration is often ad hoc, not systematically planned or embedded within existing methodological or regulatory frameworks. We outline specific strategies focusing on prospective planning, methodological transparency (such as preregistration of RWE observational studies), and robust governance frameworks to systematically integrate these complementary evidence types.
What this adds to what is known?
This paper promotes evolving beyond traditional binary perspectives on evidence types (RCT vs RWE), clarifying how innovative hybrid designs, adaptive randomized trials, and pragmatic RCTs can systematically integrate RWE to enhance generalizability without sacrificing methodological rigor. We suggest approaches for integrating RWE and randomized trials (eg, external control arms, pragmatic RCTs), addressing existing knowledge gaps and reducing redundancy or fragmented integration practices.
What is the implication, what should change now?
Stakeholders across the health-care research ecosystem, including regulatory bodies, industry, academia, and patient communities, should adopt harmonized methodological standards and robust governance frameworks for integrating randomized trials with RWE data. Most importantly, prospective planning of integrated randomized trial and RWE approaches—rather than retrospective data assembly—must become standard practice in health-care research. Immediate priorities include developing comprehensive frameworks and prospective protocols for systematic integration, enhancing methodological transparency, and ensuring robust data governance, alongside systematic evaluation of stakeholder engagement to ensure research relevance and applicability. In addition, there must be concerted efforts to develop better methodological approaches for synthesizing different data types across diverse study designs, enabling more reliable integration of evidence from heterogeneous sources.peer-reviewe
Determinants of human-machine interaction technology usage: An automated machine learning approach
No full textThe advent of Industry 4.0 technologies has reshaped modern manufacturing. Human-machine interaction (HMI) technologies are essential to this transformation, as they facilitate communication between people and machines, bridge the digital and physical worlds, improve decision-making, and increase overall productivity. However, the diffusion of these cutting-edge technologies varies greatly, possibly resulting in persistent geographical disparities over time. Moreover, our understanding of the factors determining the use of HMI technologies is still limited. Our goal is to investigate the factors that influence manufacturing firms’ use of these technologies, providing a comprehensive perspective. Combining insights provided by economic geography and innovation studies, we take a holistic approach that includes a wide range of technological, organizational, and environmental (TOE) factors. Using Automated Machine Learning (AML), we identify non-linear relationships between key predictors and the usage of HMI technology. Our analysis highlights the importance of geographical and organizational proximities in absorbing local external knowledge and coordinating long-distance knowledge pipelines alongside traditional factors influencing the rate of technology use.This work was supported by the Ministry of Science and Higher Education in Poland, grant no. 004/RID/2018/19 entitled “Economics in the face of the New Economy” financed by the Ministry of Science and Higher Education in Poland within the Regional Excellence Initiative.peer-reviewe
